---
title: "Consumption In The US"
output: 
  flexdashboard::flex_dashboard:
    theme:
      version: 4
      bootswatch: minty
      navbar-bg: "purple"
    orientation: columns
    vertical_layout: fill
    source_code: embed 
---

```{r setup, include=FALSE}
library(flexdashboard)
library(tidyverse)
library(DT)
library(plotly)
library(lubridate)
GDPdata<-read_csv("Table1_GDPdata.csv")
SeriesReport<-read_csv("SeriesReport.csv")
UnemploymentData<-read_csv("civilian-unemployment-rate.csv")
```
Brief Overview 1
===

Column {data-width=650}
---

Consumption In The US - Unemployment Analysis

In this project, I chose to explore correlations between unemployment rates and consumer spending habits in the US between 2014 and 2024. I found this topic to be interesting because I'm an Economics major. I pulled Gross Domestic Product (GDP) data, Series Report data from full-time workers, and Unemployment Rate data. I want to observe spending trends in periods of low unemployment and compare it to periods of higher unemployment rates. In this project there are significant number discrepancies in the year 2020 when the pandemic happened. 

Column {.tabset data-width=550}
---

### Graphical Displays
  - Line Graph 
  - Scatterplot
  
### Common Arguments 
List of common arguments 

  - col: a vector of colors
  - main: title for the plot
  - font: font used for text, 1=plain, 2=bold, 3=italic, 4=bold italic 
  - font.axis: font used for axis
  - cex.axis: font size for x and y axes
  - font.lab: font for x and y labels
  - cex.lab: font size for x and y labels

GDP Data
===
Column {data-width=550}
---

### <b><font size = 4><span Style = "color:blue">First 500 Observations</span></font></b>

```{r GDP}
GDP<-c(-1.4,5.3,5.0,2.0,3.6,2.5,1.6,0.7,2.3,1.3,2.9,2.2,2.0,2.3,3.2,4.6,3.3,2.1,2.5,0.6,2.5,3.4,4.8,2.8,-5.5,-28.1,35.2,4.4,5.6,6.4,3.5,7.4,-1.0,0.3,2.7,3.4,2.8,2.4,4.4,3.2,1.6,3.0,3.1,2.4)
Personal_Consumption<-c(1.4,3.9,4.0,4.7,3.1,2.8,2.8,1.6,3.1,2.0,2.8,2.1,3.1,2.0,2.7,4.4,2.9,2.2,1.9,1.3,0.6,3.5,4.6,2.7,-6.6,-30.6,41.2,5.8,9.5,14.1,3.1,4.4,1.0,2.6,1.5,1.2,4.9,1.0,2.5,3.5,1.9,2.8,3.7,4.0)
Durable_Goods<-c(6.9,15.8,7.1,8.0,7.9,8.3,5.3,2.6,5.4,3.4,10.7,6.4,3.3,5.2,10.7,15.7,5.3,2.3,2.9,0.7,-3.5,12.2,9.8,3.3,-17.2,-1.6,100.8,4.5,31.0,14.7,-24.8,8.6,0.1,-2.2,-1.9,-2.0,17.1,-0.3,4.2,2.9,-1.8,5.5,7.6,12.4)
Nondurable_Goods<-c(1.3,4.1,3.4,4.9,4.0,2.4,4.3,1.5,3.8,3.6,1.3,0.7,3.8,3.9,2.3,5.3,3.0,0.3,1.3,2.8,3.5,4.6,3.9,0.4,6.0,-12.5,31.0,2.3,10.9,14.2,0.4,2.5,-2.7,-1.2,-2.5,0.1,2.5,-0.4,3.1,3.6,-0.8,1.7,4.6,3.1)
df<-data.frame(GDP,Personal_Consumption,Durable_Goods,Nondurable_Goods)
datatable(df,rownames=FALSE,colnames=c("Total GDP","Personal Consumption GDP","GDP of Durable Goods","GDP of Nondurable Goods"))
```

Column {data-width=450}
---

### <font size = 4><span Style = "color:red">Description</span></font>

Variables used in this data set represent GDP numbers in each Quarter of the year from 2014-2024. 
Some examples of variables: 
- Gross domestic product: total GDP for that quarter
- Personal consumption expenditures: GDP of personal items for that quarter
- Durable goods: GDP of goods that don't need to be bought frequently (houses,cars)
- Non-durable goods: GDP of goods that need to be rebought frequently (toiletries, clothes)

```{r}
glimpse(GDPdata)
```

GDP Line Graph
=== 

Column {data-width=650}
---

```{r}
plot(GDP,type="o",col="blue",xlab="Number of Quarter in years 2014-2024 (ex: 1=Q1-2014 & 26=Q2-2020)",ylab="GDP Amount",xlim=c(0,44),ylim=c(-25,101))
lines(Personal_Consumption, type="o",col="orange")
lines(Durable_Goods, type="o",col="red")
lines(Nondurable_Goods,type="o",col="green")
legend("topleft",legend=c("Total GDP","Personal Consumption GDP","GDP of Durable Goods","GDP of Nondurable Goods"),col=c("blue","orange","red","green"),lty=1,pch=1)
```

### Analysis 
---

In this data representation, we see a significant decrease in Quarter 2 of 2020 (around the number 26 on the x-axis). The reason for this decrease in total GDP was because of the start of COVID-19. During this time, there was a very large drop in personal spending on services such as travel and entertainment. We can see that durable goods didn't see too much of a decrease. This is because durable goods were still being bought during this time (houses, cars, etc.) Non-durable goods saw a slight decrease as well, but not as much as personal spending. I related this to people bulk-buying toilet paper and other household goods. 

Another interesting observation is the large spike in GDP in Q3 of 2020. This is because of businesses reopening and citizens starting to resume their lives pre-pandemic. 

Post pandemic, we see a difference in consumer spending habits. 

Median Weekly Earnings Data {data-orientation=rows}
===

Column {data-width=550}
---

### <b><font size = 4><span Style = "color:blue">First 500 Observations</span></font></b>

```{r}
datatable(SeriesReport,rownames=FALSE)
```

Column {data-width=450}
---

### <font size = 4><span Style = "color:red">Description</span></font>

Variables in this data set represent usual median weekly earnings of full-time workers. These numbers are taken from an average of data from men and women. Like the rest of the data given for this project, the data is split into four quarters for each year, 2014-2024. 

```{r}
glimpse(SeriesReport)
```


Weekly Earnings Linegraph
===

Row {data-height=350}
---

```{r linegraph}
Series <- pivot_longer(SeriesReport, cols = starts_with("Qtr"), names_to = "Quarter", values_to = "Earnings")
ggplot(Series, aes(x = Year, y = Earnings, color = Quarter))+
  geom_line(size = 1) +
  geom_point() +
  labs(title = "Quarterly Trends Over Years", x = "Year", y = "Earnings")
```

### Analysis
---

As we can see, there is an overall increase in weekly earnings over the last decade. In the year 2020, we see a slight spike in the numbers, especially quarter 2. This spike is because of a mass amount of minimum-wage workers losing their jobs. This caused the average weekly earnings number to rise because data was collected from mostly full-time mandatory workers. 

A rise in weekly earnings over the last 10 years is a result of the growing economy in the US. As we see lower unemployment rates and a tight workforce, employers need to increase wages to attract and keep workers. 

Unemployment Data
===

Column {data-width=500}
---

Variables in this data set represent the percentage of total unemployment in the US in each month from years 2014-2024. This data is from working-age individuals. 

```{r}
datatable(UnemploymentData,rownames = FALSE)
glimpse(UnemploymentData)
```
Unemployment Linegraph 
===

Column {data-width=550}
---

```{r}
UnemploymentData$Month <- parse_date_time(UnemploymentData$Month, orders = "y-b")
ggplot(UnemploymentData, aes(x = Month, y = Total)) +
  geom_line(color = "lightblue", size = 1) +
  labs(title = "Civilian Unemployment Rate Over Time",
       x = "Date", y = "Unemployment Rate (%)") +
  theme_minimal()
```

Column {data-width=550}
---

### Analysis 
---

This graph of unemployment rate in the last ten years can be very helpful in analyzing the state of the economy at these times. From 2014-2020, we see that unemployment rates are slowly declining, which means that we had a lot of people in jobs during this time. But in Q2 of 2020, we see an extreme spike in the rate of unemployment. This is due to the pandemic, which caused stay-at-home orders for people in non-essential careers. As the government ordered businesses to shut down, people were laid-off of jobs. In the years 2022 and 2023, COVID-19 was beginning to become less of a prevalent issue in the US, and we are still trying to get back to pre-pandemic unemployment rates. 

GDP & Unemp. Scatterplot
===

```{r}
ScatterplotData<-read_csv("ScatterplotData.csv")
library(plotly)
attach(ScatterplotData)
```
```{r}
ggplotly(ScatterplotData %>% ggplot(aes(x=Unemp_Rate, y=GDP, colour = as.factor(Year)))+
  geom_point())
```
### Analysis
---

This graph of comparing the Unemployment Rate to GDP doesn't seem to have much correlation. This means that GDP in the US isn't very affected by the unemployment rate. There is a small cluster at around 5% of unemployment. This 5% has been the "natural rate of unemployment" in the last 10 years. The natural rate of unemployment is to account for possible slack in the labor force. At the economy's most functioning state, there will always be structural unemployment (people between jobs as the economy and labor force fluctuates) 

There are two points that differ from the rest, 2020 Q2 and Q3. Q2's unemployment rate is very high with a low GDP because of COVID. Q3's unemployment rate is generally higher and has a much larger GDP because of lifted COVID restrictions. 

GDP & Weekly Earnings Scatterplot
===

```{r}
ggplotly(ScatterplotData %>% ggplot(aes(x=Series_Report, y=GDP, colour = as.factor(Year)))+
  geom_point())
```
### Analysis 
---

As we can see, over the years weekly earnings increases relatively steadily. Once again in 2020 there are large discrepancies between the points for each quarter.  The average weekly earnings in most years are the same (vertical line of points). For years that didn't have the same amount of average weekly earnings for every quarter, this is just an effect of the changing labor force and job market. 

Boxplots by Quarter
===

Column{width=350}
---


```{r}
long_data <- ScatterplotData %>%
  pivot_longer(cols = c(Durable_Goods, Nondurable_Goods, GDP, Series_Report),
               names_to = "Category",
               values_to = "Value")
ggplot(long_data, aes(x = Quarter, y = Value, fill = Category)) +
  geom_boxplot() +
  facet_wrap(~ Category, scales = "free_y") +
  theme_minimal() +
  labs(title = "Boxplots by Quarter",
       x = "Quarter",
       y = "Value") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))
```
### Analysis
---

There isn't too much correlation between each quarter individually, but this boxplot representation is valuable because it shows us the features of Durable/Non-durable goods, GDP, and weekly earnings. 

I think Q2 has the most recognizably different spread of data. 

It seems that across all 4 data sets, Q4 has the smallest range. 

Overall Findings, Limitations, and Further Research
===

Overall, we see that during periods of low unemployment or a natural rate of unemployment, there is likely a higher GDP because people are earning enough to spend money on personal items, whether that be durable or non-durable goods. 

In periods of very high unemployment, we see a lower GDP and not too much a discrepancy between weekly earnings. 

As a general trend, GDP varies because of outside events, so it is likely always changing. But large economic events such as COVID cause large differences in GDP. 

Weekly earnings (series report data) doesn't see much of an impact from outside events. But we can see an increase over the past 10 years as a steady change. 

Limitations: 
Difficulties for this project were the fact that there sometimes isn't much variation in the economy, so data representations are "boring" and don't show many key features. The economy is always changing when there are no sudden shocks. I chose to focus on differences in data in the year 2020 because it was the most notable

Further Research: 
I think it would be interesting to look at the economy from the last quarter because there have been signs of a recession. Economists say we are at a 40-60% chance of a recession due to the tariffs and other economic decisions being made currently.